JPWO2007135884A1 - Ultrasonic diagnostic apparatus and ultrasonic diagnostic method - Google Patents

Abstract

In the ultrasonic diagnostic apparatus, a selection unit that selects a part for detecting the position of the boundary of the organ of the subject on the ultrasonic image displayed on the display unit, and a predetermined interval on the ultrasonic image. Boundary extraction filter setting means for setting a boundary extraction filter composed of the two regions obtained, and analyzing pixel data in the boundary extraction filter set by the boundary extraction filter setting means in the vicinity of the part selected by the selection means Accordingly, a boundary position detecting means for detecting the position of the boundary is provided, and the boundary position detected by the boundary position detecting means is displayed on the display means under the control of the control means.

Description

  The present invention relates to an ultrasonic diagnostic apparatus and an ultrasonic diagnostic method, and more particularly to an ultrasonic diagnostic apparatus and an ultrasonic diagnostic method that can accurately depict a boundary (contour) of a target tissue such as an organ.

  In ultrasonic diagnosis using an ultrasonic diagnostic apparatus, boundary (contour) information of an organ or the like of interest is useful diagnostic information. For example, in an organ such as the heart, the boundary (contour) of the left ventricle is drawn, the area of the region surrounded by the boundary (contour) is obtained, or the volume of the left ventricle is estimated from the boundary (contour) Is considered useful in diagnosis.

  As a technique for extracting and displaying a tissue boundary (contour), for example, there is a conventional technique described in Patent Document 1.

JP-A-8-206117.

  According to the prior art described in Patent Document 1, for example, a maximum point of a scalar amount representing a gradient of an image signal at each point is obtained from a tomographic image, and a boundary (contour) of a tissue is drawn from the maximum point.

However, the present inventors have found the following problems as a result of examining the above prior art.
That is, noise called speckle noise is mixed in the ultrasonic image. This speckle noise is thought to appear when the scattered waves from the reflectors in the living tissue that are sufficiently smaller than the wavelength of the ultrasonic waves are generated and interfered with in various phases (known techniques related to speckle noise). (See Patent Document 2).

JP-A-9-94248.

  According to the prior art described in Patent Document 1 above, the boundary (outline) of the organ is extracted based on the maximum point of the scalar amount representing the gradient, so that the influence of blur caused by speckle noise is taken into consideration. Was not. For this reason, the organ boundary (contour) has a problem that a region having a high pixel density protrudes to a region having a low pixel density by the blur width and the boundary is extracted. For example, when trying to extract the boundary (contour) of the left ventricle of the heart, the heart wall is extracted inward by the blur width, which causes a problem that the size of the left ventricle is smaller than the actual size.

According to the present invention, an ultrasound probe that transmits and receives ultrasound to and from a subject, and an ultrasound image connected to the ultrasound probe and based on an ultrasound signal obtained by the ultrasound probe. Connected to the ultrasonic probe and the image generation means, and controls them, and connected to the image generation means and the control section, and controlled by the control section In an ultrasonic diagnostic apparatus comprising display means for displaying an ultrasonic image generated by the image generation means based on:
On the image displayed on the display means, a selection means for selecting a part for detecting the position of the boundary of the organ of the subject, and two regions spaced by a predetermined interval on the ultrasound image Analyzing the pixel data in the boundary extraction filter set by the boundary extraction filter setting means in the vicinity of the part selected by the selection means; An ultrasonic diagnostic apparatus comprising: boundary position detection means for detecting a position of the boundary; and the boundary position detected by the boundary position detection means is displayed on the display means under the control of the control means. Provided.

Further, according to the present invention, in the ultrasonic diagnostic method capable of extracting the position of the boundary of the organ appearing on the ultrasonic image,
(1) On the ultrasonic image, designating a part for extracting the boundary;
(2) calculating the blur width of the image of the vicinity of the site set in the step (1);
(3) a step of setting two regions having an interval according to the blur width calculated in the step (2) as a boundary extraction filter;
(4) While changing the position and / or inclination of the boundary extraction filter set in the step (3), analyzing the pixel values in the two regions to obtain the boundary strength;
(5) in the step (4), the step of extracting the position of the boundary by obtaining the position and / or the slope at which the boundary strength is maximum or a predetermined value or more;
(6) Based on the boundary position obtained in the step (5), comprising the step of calculating the area of the region surrounded by the boundary or the volume of the organ representing the region surrounded by the boundary An ultrasonic diagnostic method is provided.

  An object of the present invention is to extract a contour with high accuracy in consideration of a blur width appearing on an image in an ultrasonic diagnostic apparatus that extracts an organ boundary (contour) using an ultrasonic image.

1 is a block diagram showing the overall configuration of an ultrasonic diagnostic apparatus according to Embodiment 1 of the present invention. FIG. 3 is a diagram showing the inside of the control unit 3. FIG. 6 is a diagram illustrating an example of image data that is a target of boundary extraction processing according to the first embodiment. The figure which compared and showed the boundary position 18 calculated by the conventional method, and the true boundary position 19. FIG. 3 is a flowchart of boundary extraction processing according to the first embodiment. The figure which shows what kind of speckle arises according to the kind of ultrasonic probe to be used. Filter shape of the boundary extraction filter with two areas. The figure explaining how the boundary of two area | regions 28-1 and 28-2 is blurred by the speckle which appears on an image. The figure which shows the boundary extraction filter produced when extracting the boundary formed in the closed area | region correctly. The figure for calculating the statistical value for calculating boundary strength using a degree of separation. The figure which shows the boundary extraction filter set by bending the area | region of the space | gap pinched | interposed into two area | regions according to the assumed boundary being bent. The figure which shows the boundary extraction filter for extracting a boundary position suitably from the four-chamber section image or short-axis section image of the heart. 10 is a flowchart of boundary extraction processing according to the second embodiment. The figure which shows the example in the case of changing the space | interval of a filter area | region, without changing the shape of two filter area | regions. The figure which shows the example in the case of changing the shape of two filter area | regions.

Explanation of symbols

  7 Boundary extraction part designation means, 8 Boundary extraction calculation means, 9 Organ calculation means, 10 Image blur width calculation means, 11 Filter shape creation / deformation means, 12 Boundary strength calculation means, 13 Boundary position detection means

  Hereinafter, an ultrasonic diagnostic apparatus according to an embodiment of the present invention will be described based on the drawings.

  FIG. 1 is a block diagram showing the overall configuration of the ultrasonic diagnostic apparatus according to Embodiment 1 of the present invention.

  According to FIG. 1, the ultrasound diagnostic apparatus generates an ultrasound image based on an ultrasound probe 1 that transmits and receives ultrasound and an ultrasound signal that is connected to the ultrasound probe 1 and received by the ultrasound probe. An image generation unit 2, a control unit 3 such as a CPU (Central Processing Unit) that is connected to each component of the ultrasonic diagnostic apparatus and performs operation control and arithmetic processing of each component, and ultrasonic diagnosis An operation unit 4 that is connected to each component of the apparatus and that allows an operator such as a medical worker to operate the ultrasonic diagnostic apparatus using an input device (keyboard, mouse, trackball, touch panel, etc.), and an ultrasonic diagnostic apparatus A storage unit 5 that stores image data, a program, and the like connected to each of the components, and a display unit such as a CRT or a liquid crystal display device that is connected to each component of the ultrasonic diagnostic apparatus and displays images, measurement results, and the like It consists of six.

  The ultrasound probe 1 transmits and receives ultrasound to and from the living body of a subject, and is a linear type in which transducers are arranged in a straight line, and the angle of the beam by driving the transducer with a time difference There are a sector type that can change the frequency, and a convex type that performs scanning while shifting transducers by arranging transducers in a convex shape. The ultrasonic probe 1 converts ultrasonic waves (ultrasonic echoes) reflected and returned from the living body of the subject into electrical signals and sends them to the image generation unit 2.

  The image generation unit 2 generates a B-mode image using the signal received by the ultrasonic probe 1 and converted into an electrical signal as an input signal. The input signal is converted into a B-mode image through a phasing adder, a logarithmic amplifier, an envelope detector, an A / D converter, and a scan converter in the image generation unit 2.

  The control unit 3 is executed by loading a control program for the ultrasonic diagnostic apparatus stored in the storage unit 5 or the like. The control unit 3 instructs each component of the ultrasonic diagnostic apparatus to perform an operation, and performs timing control and arithmetic processing.

  The operation unit 4 is an input device such as a keyboard, a mouse, a trackball, and a touch panel on the ultrasonic diagnostic apparatus, and is used by a diagnostician such as a medical staff for image quality adjustment, measurement instruction, information input, and the like.

  The storage unit 5 is a device that stores image data, a control program, and the like, and is a hard disk, a general-purpose memory, a frame memory, or the like. The image data stored in the storage unit 5 is an acquired B-mode image or a file in an image format that can be displayed on a general PC.

  The display unit 6 is a CRT or a liquid crystal display device that displays image data, measurement values, and an image obtained by graphing the measurement values on a screen.

  FIG. 2 shows the inside of the control unit 3 according to Embodiment 1 of the present invention. According to FIG. 2, in the control unit 3 of the present invention, the boundary extraction part designating means 7 for designating which part of the ultrasound image the boundary of the organ is to be extracted, and the boundary extraction part designating means 7 Based on the boundary extraction calculation means 8 for extracting the boundary of the organ in the vicinity of the part specified by 7 and the boundary extracted by the boundary extraction calculation means 8, various physical quantities, that is, distances such as the size of the organ Organ measuring means 9 for calculating an area on the organ image, an estimated value of the volume of the organ, and the like is provided.

  The boundary extraction part designating means 7 is a means for the operator to designate the vicinity near the boundary extraction target on the image displayed on the screen of the display unit 6 by the input device. The boundary extraction part designating means 7 may automatically determine the boundary extraction part by performing signal processing on the pixel value of the acquired image data.

  The boundary extraction calculation means 8 calculates the image blur width of the part of the image to be extracted and indexed by the boundary extraction part designation means 7, and selects an appropriate filter in consideration of the image blur width. This is a means for detecting a position, and comprises an image blur width calculating means 10, a filter shape creating / deforming means 11, a boundary strength calculating means 12, and a boundary position detecting means 13.

  First of all, the image blur width calculating means 10 is a means for calculating the size of the speckle using the pixel values in the vicinity of the boundary extraction portion designated by the boundary extraction portion designation means 7. A concentration co-occurrence matrix or an autocorrelation function is used as a method for calculating the size of speckle.

  Next, the filter shape creation / deformation means 11 creates and transforms the boundary extraction filter. For example, the filter shape creation / deformation means 11 creates a filter composed of two regions separated by a distance based on the speckle size calculated by the image blur width calculation means 10. Further, if the shape of the boundary to be extracted is known, the filter shape creation / deformation means 11 can also change the shape of the filter according to the shape of the boundary of this shape.

  Next, the boundary strength calculating unit 12 is configured to change the pixel values in the two regions at each position and / or inclination while moving the boundary extraction filter created by the filter shape / deformation unit 11 at an arbitrary position and / or inclination. For example, the boundary strength is calculated by calculating the degree of separation described later.

  Next, the boundary position detection means 13 moves the position and / or inclination of the boundary extraction filter while scanning the boundary extraction filter whose boundary intensity calculated by the boundary intensity calculation means 12 is a maximum value or a predetermined value or more. Detect position and / or tilt. Based on the detected position of the boundary extraction filter, the coordinate value of the boundary position to be extracted is obtained.

  Further, the organ measuring means 9 calculates various physical quantities related to the organ from which the boundary is extracted, for example, distance, area, volume, and the like, using the coordinate value of the extracted boundary position. For example, the organ measuring means 9 calculates a physical quantity such as the size of the tumor of the affected area that is the target region with high accuracy.

  Next, FIG. 3 shows an example of image data to be subjected to boundary extraction processing in the first embodiment. In FIG. 3, 14 is image data, 15 is a beam direction, 16 is a beam depth, 17-1 to 17-3 are speckles, and 18 is a boundary position. In 17-1 to 17-3, speckles become smaller in size in the order of 17-1 → 17-2 → 17-3, and the depth becomes shallower. That is, the speckle size increases as the depth increases.

  Next, FIG. 4 shows a comparison between the boundary position 18 calculated by the conventional method and the true boundary position 19. According to this, it can be seen that a large error occurs between the boundary position 18 calculated by the conventional method and the true boundary position 19.

Next, a flowchart of boundary extraction processing according to the first embodiment will be described with reference to FIG.
First, in the ultrasonic diagnostic apparatus, the ultrasonic probe 1 and the image generation unit 2 acquire image data 14 obtained by imaging a patient organ or the like, and start boundary extraction for a target region such as the organ.
(Step 20)
First, the ultrasonic diagnostic apparatus selects and inputs a part to be subjected to boundary extraction manually or automatically by the boundary extraction part designating means 7.
(Step 21)
Next, the ultrasonic diagnostic apparatus calculates the blur width on the image at the boundary extraction site specified in step 20 by the image blur width calculation means 10. Specifically, texture analysis is performed on the pixel value data in the vicinity of the boundary extraction region, speckles appearing on the image are approximated by an elliptical shape, and the width of the approximated ellipse (long or short axis length) is calculated. Half the distance is calculated as the blur width.

  As an example of the texture analysis here, there is a method of obtaining a density co-occurrence matrix or an autocorrelation function. (For example, O. Bassel, et al: “TEXTURE ANALYSIS OF ULTRASONIC IMAGES OF THE PROSTATE BY MEANS OF CO-OCCURRENCA MATRICS.” ULTRAONIC IMAGING 15, 218-237 (1993), BJOosterveld et al: “TEXTURE OF B-MODE. ECHOGRAMS: 3-D SIMULATIONS AND EXPERIMENTS OF THE EFFECTS OF DIFFRACTION AND SCATTERER DENSITY ”ULTRASONIC IMAGING 7,142-160 (1985)) By performing such a texture analysis, the blur width (or speckle width of an ultrasonic image) For example, even when a parameter related to an ultrasonic sound field that affects a half distance) is unknown, the blur width can be calculated using only information obtained from the received signal.

  Here, FIG. 6 shows what speckles are generated according to the type of ultrasonic probe used. Fig. 6 (a) shows an example of a linear ultrasonic probe 25-1.Speckle 17 when approximated by an ellipse has its major and minor axes aligned in the horizontal and vertical directions on the screen. I'm doing it. Therefore, even when the speckle size is extracted by approximation with an ellipse, the blur width is calculated so that the directions of the major axis and the minor axis coincide with the horizontal direction and the vertical direction on the screen.

On the other hand, FIG. 6 (b) is an example of the sector-type ultrasonic probe 25-2, and the speckle 17 when approximated by an ellipse has one of its major axis and minor axis pointing in an oblique direction on the screen. And is in a direction to transmit and receive ultrasonic waves. Therefore, even when extracting the speckle size by approximating with an ellipse, the blur width is set so that one of the long axis and the short axis coincides with the oblique direction on the screen, that is, the direction in which ultrasonic waves are transmitted and received. calculate.
(Step 22)
Next, the filter shape / deformation means 11 creates a boundary extraction filter having two regions as shown in 26-1 and 26-2 in FIGS. 7 (a) and 7 (b). The boundary extraction filters (27-1 and 27-2) in this step consist of two areas 26-1 and 26-2, and the distance between the opposite sides 28-1 and 28-2 of the two areas 29 is equal to the blur width obtained in step 21.

  Here, how the boundary between the two regions 30-1 and 30-2 is blurred by speckle appearing on the image will be described with reference to FIG. First, FIG. 8 (a) shows two regions and a true boundary position 19 located between them. Next, FIG. 8 (b) shows one of the speckles 17 appearing on the image, and 17-4 is the width of the speckle in the horizontal direction of the drawing. Next, FIG. 8 (c) is a profile of an ultrasonic image and shows a pixel distribution on a line segment crossing a true boundary position. According to this, it can be seen that the region (30-2) having a high pixel value protrudes to the left in the drawing because the speckle has a width in the horizontal direction of the drawing. When the boundary extraction described in Patent Document 1 is directly performed using the profile as shown in FIG. 8 (c), the boundary position is extracted at a position different from the original position as shown at 18. Therefore, in order to detect the true boundary position in this step, a boundary extraction filter is used so that the position where the parameter for detecting the boundary position (intensity intensity described later) is the maximum becomes the true boundary position 19. Must be created.

  7 (a) and (b), the bottom side of FIG. 7 (a) is lower when the pixel value is higher on the right side of FIG. 7, and the lower side of FIG. When the pixel value is high, the profile is similar to that shown in FIG. 8 (c), and the boundary extraction filter created in this step faces each other in the two regions 26-1 and 26-2. The distance between the sides 28-1 and 28-2 is equal to the blur width of the profile on the lower side of FIGS. 7 (a) and 7 (b). This blur width is obtained in step 21, and is set to, for example, 1/2 of the speckle width in FIG. 8 (b).

  FIG. 9 shows a boundary extraction filter that is created when the boundary formed in the closed region is accurately extracted. More specifically, as shown in the pixel value profile on the lower side of FIG. 9, the surrounding high pixel value regions (31-1 and 31-2) and the high pixel value region (31-1 And the boundary position with the low pixel value region (31-3) surrounded by 31-2).

In this case, as shown in the upper side of FIG. 9, the inner circular region 32-1 and the outer ring-shaped region 32-2 are spaced from each other by a gap 33 having a blur width calculated in step 21. To create a boundary extraction filter.
(Step 23)
Next, the boundary strength calculation means 12 performs boundary strength calculation by scanning the boundary creation filter created in the process of step 22 within the image. In this embodiment, for example, the boundary strength calculation will be described using a degree of separation that is a ratio of the variance between two classes and the total variance in two regions described later. However, the variance between classes and the total variance in the two areas are as described in IEICE Transactions Vol.J63-D No.4, p349-p356. After the data is averaged in each region, the variance value is obtained. The total variance is obtained by using the pixel data in the two regions as they are. Hereinafter, details of the calculation of the degree of separation in the present embodiment will be described with reference to FIG. 10 and equations (1) to (3).

  FIG. 10 is a diagram for calculating a statistical value for calculating the boundary strength using the degree of separation, in which two regions 26-1 and 26-2 and a blur width 29 are shown.

As shown in FIG. 10, the numbers of pixels in regions 26-1 and 26-2 are N ₁ and N ₂ , and the average luminance of each region is μ ₁ , μ ₂ , and regions 26-1 and 26-2 are combined. When the average luminance value of the region is μ and the luminance value of each pixel is defined as Pi (1 ≦ i ≦ N ₁ , 1 ≦ i ≦ N ₂ ), the degree of separation η is expressed by the following equations (1) to (3). expressed.

式(2)

式(3)

(ステップ24)
図7の26-1、26-2で示されたような境界抽出フィルタを画像上でいろいろな位置及び/あるいは傾きに動かせて走査させながら、逐次式(1)〜(3)を計算して、分離度(境界強度)がどのように変化するかを計算する。そして、図8(d)で示されたような境界強度の分布を求めて、境界強度が最大となる位置を真の境界位置19として求めて検出する。Formula (1)

Formula (2)

Formula (3)

(Step 24)
While calculating the boundary extraction filter as shown in 26-1, 26-2 in Fig. 7 by moving it to various positions and / or inclinations on the image, calculate the sequential formulas (1) to (3). Calculate how the degree of separation (boundary strength) changes. Then, the distribution of the boundary strength as shown in FIG. 8D is obtained, and the position where the boundary strength is maximum is obtained as the true boundary position 19 and detected.

  However, although the boundary extraction filters shown in FIGS. 7 (a) and (b) have the same shape, the direction of the region with a high pixel value differs between FIGS. 7 (a) and 7 (b). The boundary position to be extracted is varied depending on the case. For example, in the case of FIG. 7 (a), since the pixel value is high on the right side of the drawing, the edge of the region 26-2 set on the higher pixel value side of the two regions constituting the boundary extraction filter Then, a position closer to the other region (26-1) is extracted as a reference point (33) for boundary position detection. In the case of FIG. 7B, since the pixel value is high on the left side of the drawing, the edge of the region 26-1 set on the higher pixel value side of the two regions constituting the boundary extraction filter. Then, a position closer to the other region (26-2) is extracted as a reference point (34) for detecting the true boundary position 19.

  In addition, when using the boundary extraction filter created when accurately extracting the boundary formed in the closed region as shown in FIG. 9, the region on the higher pixel value side (in the example of FIG. 9, The position inside the outer ring-shaped region 32-2) is extracted as a reference point (35) for detecting the true boundary position 19.

  When the boundary extraction processing of the target region is completed through the above steps 20 to 24, the obtained boundary position (boundary information) is displayed on the display unit 6, and the target whose boundary is extracted by the organ measuring means 9 A physical quantity such as the size of the target area is calculated using the boundary information of the area, and the calculated value is displayed and recorded.

  According to the ultrasonic diagnostic apparatus according to the first embodiment, the blur width is calculated from the ultrasonic image, the shape and size of the boundary extraction filter are set based on the blur width, and the boundary position of the target region is extracted. Therefore, the boundary extraction of the target area can be performed with high accuracy. More specifically, the boundary extraction filter used in the above embodiment includes two regions, and the width of the gap sandwiched between the two regions is equal to the blur width calculated from the ultrasonic image. In consideration of (that is, a distance corresponding to, for example, 1/2 of the size of speckle appearing on the ultrasonic image), the true boundary position can be extracted with high accuracy. In the conventional boundary extraction method, the boundary extraction is performed assuming that the pixel value changes stepwise at the boundary. Therefore, the boundary position detection accuracy increases as the boundary appearing on the actual image becomes dull due to the influence of speckle or the like. It was getting worse. In the present invention, since the boundary is extracted assuming that the boundary is dull due to the influence of speckle or the like, it is possible to always extract the boundary position with the same accuracy as when the boundary is not blunt on the image. Become. Then, by using the calculated coordinates of the true boundary position, a physical quantity such as the size of the target region can be accurately calculated, and accurate ultrasonic diagnosis can be performed using the value.

  Note that the speckle size (for example, an amount corresponding to the blur width) increases as the image acquisition depth in the subject increases. Therefore, it is considered desirable to change the interval between the two regions constituting the boundary extraction filter according to the image acquisition depth.

  Further, it may be desirable to set the boundary extraction filter to be set so as to reflect the boundary shape of the target part. A boundary extraction filter having the shape shown in FIG. 7 or FIG. 9 may be used. However, when the boundary assumed in advance is bent or when a specific part of the heart is to be extracted, the boundary extraction filter is adjusted accordingly. A boundary extraction filter having a shape as shown in FIGS. 11 and 12 may be used. In FIG. 11, 36-1 and 36-2 are two regions that constitute the boundary extraction filter, respectively, but the void region sandwiched between the two regions corresponds to the assumed boundary being bent. Bent and set.

  In addition, the figure shown in FIG. 12 (a) is a four-chamber cut-off image of the heart, but each filter shown in 37-1 to 37-5 is a location of the four-chamber cut-out image of the heart. This is for suitably extracting the boundary positions for 38-1 to 38-5. In addition, the figure shown in FIG. 12 (b) is a short-axis cut image of the heart, but each filter shown in 39-1 to 39-2 has different locations of the short-axis image of the heart. The boundary position for 40-1 to 40-2 is preferably extracted. In this way, when extracting a characteristic part of the heart, a filter having a shape of a structure such as an annulus or a papilla having a characteristic shape may be created. If the contour of the heart can be suitably extracted using the boundary extraction filter as shown in FIG. 12, the contour of the lumen is blurred inward when the volume of the heart cavity is measured. It is possible to prevent the volume value from being underestimated.

  Next, Embodiment 2 of the present invention will be described with reference to FIGS. The second embodiment uses the boundary extraction filter according to the present invention to search the boundary when the information about the blur width of the image or the shape of the boundary to be extracted is unknown or when calculation is not possible for some reason. It is an example which extracts a position.

For example, if the speckle distribution in the target area of the acquired image is not uniform, the speckle size cannot be calculated correctly because the statistical properties of the speckle itself are not reflected in the density co-occurrence matrix or autocorrelation function. In such a case, the boundary position can be searched according to the flowchart shown in FIG. Hereinafter, each step of the flowchart of FIG. 13 will be described in order.
(Step 40)
First, the ultrasonic diagnostic apparatus designates the boundary extraction part manually or automatically by the boundary extraction part designation means 7.
(Step 41)
Next, since the blur width or shape on the ultrasonic image is unknown in this embodiment, the filter shape creation / deformation means 11 creates a boundary extraction filter having an appropriate initial shape composed of two regions.
(Step 42)
Next, while changing any of the filter area interval, shape, position, or inclination of the boundary extraction filter that consists of the two areas set in step 41, the boundary strength is sequentially calculated by the method described in step 23, and the distribution is obtained. Ask for.

  For example, FIG. 14 is a diagram illustrating an example of changing the interval between the filter regions without changing the shape of the two filter regions. In this step, when the boundary strength is sequentially calculated while changing the filter area interval, the filter area interval is changed from 44-1 to 44-2, and the boundary extraction filter is also changed from 45-1 to 45-2. Change as follows.

FIG. 15 is a diagram illustrating an example in which the shapes of the two filter regions are changed. In the example of FIG. 15, the boundary extraction most suitable for the bending condition of the boundary line to be extracted is performed by deforming the shape of the opposite sides of the two filter areas in the order of 46-1 to 46-2, 46-3. The filter can be searched.
(Step 43)
In step 40, the boundary strength is calculated while changing any one of the filter region interval, shape, position, and inclination of the boundary extraction filter, and the filter region interval, shape, and position that maximize the boundary strength are obtained. When the filter area interval matches the blur width of the target area of the ultrasound image, the shape matches the shape of the boundary to be extracted, and when the position matches the position of the boundary to be extracted, the boundary strength is maximized. Extract the boundary position.

  When changing the filter region interval, if the filter region interval is changed from a smaller one to a larger one, the boundary strength gradually increases until the size of the filter region interval matches the blur width. Then, the boundary strength no longer increases. Therefore, when calculating the boundary strength while changing the filter area interval, change the boundary strength from smaller to larger, search for the interval where the boundary strength value begins to saturate at the maximum value, and detect the boundary position It is thought that it should be done.

  According to the present embodiment, even when the information about the blur width of the image or the shape of the boundary to be extracted is unknown or cannot be calculated for some reason, the boundary extraction filter, the filter area interval, the shape, the position The true boundary position can be detected by scanning the image while changing any of the above and searching for a position where the boundary intensity is maximum.

  Further, it is possible to accurately calculate a physical quantity such as the size of the target region using the calculated coordinates of the true boundary position, and perform an accurate ultrasonic diagnosis using the value.

  The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention. For example, the present invention can be applied not only to an ultrasonic diagnostic apparatus but also to image measurement performed offline on an electronic computer such as a personal computer. In FIG. 7, the boundary extraction filter is composed of two rectangular regions, but may be a circular region. Further, the size of the two regions may be arbitrary, and may be a region composed of a small number of pixels. In addition, the half distance of the ellipse width (long axis or short axis length) approximating the speckle obtained in step 21 was used as the blur width, and used in step 22 to create the boundary extraction filter. It goes without saying that any distance other than half may be calculated as the blur width according to the nature of the speckle. The parameter used as the boundary strength may be the degree of separation specified by Equations (1) to (3), but it is an arbitrary value that indicates how much the value differs between the image data contained in the two regions. Needless to say, it may be an index or an index based on another calculation method.

Claims (20)

An ultrasonic probe that transmits / receives ultrasonic waves to / from a subject, and an image generation unit that is connected to the ultrasonic probe and generates an ultrasonic image based on an ultrasonic signal obtained by the ultrasonic probe A control unit connected to and controlling the ultrasonic probe and the image generation unit; and connected to the image generation unit and the control unit, and based on control by the control unit, the image generation unit In the ultrasonic diagnostic apparatus provided with a display means for displaying the ultrasonic image generated in
On the image displayed on the display means, a selection means for selecting a part for detecting the position of the boundary of the organ of the subject, and two regions spaced by a predetermined interval on the ultrasound image Analyzing the pixel data in the boundary extraction filter set by the boundary extraction filter setting means in the vicinity of the part selected by the selection means; An ultrasonic diagnostic apparatus comprising boundary position detection means for detecting a position of the boundary, wherein the boundary position detected by the boundary position detection means is displayed on the display means under the control of the control means.   The boundary extraction filter set by the boundary extraction filter setting means calculates the boundary strength by analyzing the pixel values in the two regions while changing the position and / or inclination near the part selected by the selection means. Boundary strength calculation means is provided, and the boundary position detection means detects the position of the boundary by obtaining the position and / or inclination of the boundary extraction filter at which the boundary strength becomes maximum or a predetermined value or more. The ultrasonic diagnostic apparatus according to claim 1, wherein 3. The ultrasonic diagnostic apparatus according to claim 2, wherein an index representing how much the value is different between image data included in two regions is used as the boundary strength.   The boundary strength is obtained by averaging the pixel data in the two regions in each region and then calculating the variance value between the classes, using the pixel data in the two regions as they are, and the variance value. 3. The ultrasonic diagnostic apparatus according to claim 2, wherein the degree of separation is obtained by dividing by the total dispersion obtained by calculating.   The boundary extraction filter set by the boundary extraction filter setting means can change the interval, shape, position or inclination of the two regions, and the boundary strength calculation means can determine the interval, shape, position of the two regions. Alternatively, the boundary strength is sequentially calculated while changing the inclination, and the boundary position detection means detects the position of the boundary based on the interval, shape, position, or inclination of the two regions where the boundary strength is maximum. 3. The ultrasonic diagnostic apparatus according to claim 2, wherein   The two regions are composed of two regions of a pixel value high side and a pixel value low side, and the boundary position detection means is information on the position and / or inclination of the boundary extraction filter that maximizes the boundary strength. The edge that faces the low pixel value side of the periphery surrounding the high pixel value region at the position is defined as the boundary position based on Ultrasound diagnostic equipment.   On the basis of the boundary position obtained by the boundary position extracting means, an organ measuring means for calculating the area of the region surrounded by the boundary or the volume of the organ representing the region surrounded by the boundary is provided. The ultrasonic diagnostic apparatus according to claim 1.   A blur width calculating unit that calculates a blur width on the ultrasonic image at the part selected by the selecting unit; and the boundary extraction filter setting unit determines an interval between the two regions by the blur width calculating unit. 2. The ultrasonic diagnostic apparatus according to claim 1, wherein the ultrasonic diagnostic apparatus is determined based on the calculated blur width.   The two regions are rectangular in shape, and the boundary extraction filter setting unit sets a distance between opposite sides of the two regions to be equal to the blur width. The ultrasonic diagnostic apparatus according to range 8.   The two regions are composed of two regions of an outer ring shape and an inner circular shape, and the boundary extraction filter setting means sets the width of the gap between the two regions equal to the blur width. 9. The ultrasonic diagnostic apparatus according to claim 8, wherein the ultrasonic diagnostic apparatus is set as follows.   9. The ultrasonic diagnostic apparatus according to claim 8, wherein the shape of the shape of the two regions by the boundary extraction filter setting unit reflects a boundary shape of a target part from which a boundary is to be extracted. .   The blur width calculation means obtains the size of the ellipse at each depth according to the depth of the ultrasonic transmission / reception direction on the ultrasound image and the size of the speckle varies. 9. The ultrasonic diagnosis according to claim 8, wherein the boundary extraction filter setting means sets the boundary extraction filters having different intervals according to a change in speckle size according to the depth. apparatus.   9. The ultrasonic diagnostic apparatus according to claim 8, wherein each of the two regions has a circular shape, and is set at a distance corresponding to the blur width.   9. The ultrasonic diagnostic apparatus according to claim 8, wherein the blur width calculating unit calculates a blur width on the image by texture analysis.   15. The ultrasonic diagnostic apparatus according to claim 14, wherein the texture analysis is performed by calculating a density co-occurrence matrix and an autocorrelation function using image data of the ultrasonic image.   15. The super width according to claim 14, wherein the blur width calculating means calculates a blur width by approximating speckles appearing on the ultrasonic image with an elliptical shape and obtaining a size of the ellipse. Ultrasonic diagnostic equipment.   17. The ultrasonic diagnostic apparatus according to claim 16, wherein a distance half the length of a short axis or a long axis of an ellipse approximating the speckle is calculated as the blur width.   15. The blur width calculating unit performs the texture analysis in consideration of an orientation on the ultrasonic image of the speckle that changes depending on a type of the ultrasonic probe. The ultrasonic diagnostic apparatus as described.   When the ultrasonic probe is a linear type, the blur width calculation means determines the orientation of the minor axis or major axis of the ellipse for approximating the speckle in the vertical direction or horizontal direction on the screen of the display means. 17. The ultrasonic diagnostic apparatus according to claim 16, wherein when the ultrasonic probe is a sector type, a direction of the ultrasonic beam is a short axis or a long axis of the ellipse. . In the ultrasonic diagnostic method capable of extracting the position of the boundary of the organ appearing on the ultrasonic image,
(1) On the ultrasonic image, designating a part for extracting the boundary;
(2) calculating the blur width of the image of the vicinity of the site set in the step (1);
(3) a step of setting two regions having an interval according to the blur width calculated in the step (2) as a boundary extraction filter;
(4) While changing the position and / or inclination of the boundary extraction filter set in the step (3), analyzing the pixel values in the two regions to obtain the boundary strength;
(5) In the step (4), the step of extracting the position of the boundary by obtaining the position and / or inclination at which the boundary strength is maximized;
(6) Based on the boundary position obtained in the step (5), comprising the step of calculating the area of the region surrounded by the boundary or the volume of the organ representing the region surrounded by the boundary And an ultrasonic diagnostic method.

Images